/**
* Copyright 2011 Pablo Mendes, Max Jakob
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.dbpedia.spotlight.lucene.similarity;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.lucene.index.IndexReader;
import org.apache.lucene.index.Term;
import org.apache.lucene.index.TermDocs;
import org.apache.lucene.search.DocIdSet;
import org.apache.lucene.search.Query;
import org.apache.lucene.util.OpenBitSet;
import org.dbpedia.spotlight.lucene.LuceneManager;
import org.dbpedia.spotlight.model.SpotlightConfiguration;
import org.dbpedia.spotlight.model.SurfaceForm;
import org.dbpedia.spotlight.model.Text;
import java.io.IOException;
import java.util.*;
/**
*
* The strategy here is to keep as many cached bitsets as possible,
* @author pablomendes
*/
public abstract class TermCache {
Log LOG = LogFactory.getLog(TermCache.class);
long mMaxCacheSize;
LuceneManager mLuceneManager;
public TermCache(LuceneManager mgr, long maxCacheSize) {
setMaxCacheSize(maxCacheSize);
this.mLuceneManager = mgr;
}
public TermCache(IndexReader reader) throws IOException {
setMaxCacheSize(reader.getUniqueTermCount());
}
/**
* It is faster to do a get(term)!=null, because in our use case we always need to get when checking if contains.
* @param term
* @return
*/
@Deprecated
public abstract boolean containsKey(Term term);
public abstract OpenBitSet put(Term term, OpenBitSet openBitSet);
public abstract OpenBitSet get(Term term);
/**
* Retrieves from the cache the DocIdSet for documents containing the context term at hand.
* If DocIdSet is not cached, creates it by reading from the index (createDocIdSet).
* @param reader
* @param term
* @return
* @throws java.io.IOException
* @author pablomendes
*/
private DocIdSet getDocIdSet(IndexReader reader, Term term) throws IOException
{
OpenBitSet contextTermDocIdSet = get(term);
if (contextTermDocIdSet == null) {
contextTermDocIdSet = createDocIdSet(reader, term);
put(term, contextTermDocIdSet);
}
OpenBitSet result = (OpenBitSet) contextTermDocIdSet.clone(); // if the outside process modifies it, our copy remains intact
return result;
}
/* (non-Javadoc)
* adapted from org.apache.lucene.search.Filter#getDocIdSet(org.apache.lucene.index.IndexReader)
* //TODO is there a faster way to do this?
*/
public OpenBitSet createDocIdSet(IndexReader reader, Term term) throws IOException
{
OpenBitSet result=new OpenBitSet(reader.maxDoc());
TermDocs td = reader.termDocs();
int c = 0;
td.seek(term);
while (td.next())
{
c++;
result.set(td.doc());
}
return result;
}
public long cardinality(IndexReader reader, Term surfaceFormTerm) throws IOException {
OpenBitSet surfaceFormDocIdSet = (OpenBitSet) getDocIdSet(reader, surfaceFormTerm);
return surfaceFormDocIdSet.cardinality();
}
/**
* Returns the number of documents containing the surface form contextTerm AND the contextTerm provided.
* @param reader
* @param contextTerm
* @return
* @throws java.io.IOException
*/
public long cardinality(IndexReader reader, Term surfaceFormTerm, Term contextTerm) throws IOException {
OpenBitSet contextDocIdSet = (OpenBitSet) getDocIdSet(reader, contextTerm);
OpenBitSet surfaceFormDocIdSet = (OpenBitSet) getDocIdSet(reader, surfaceFormTerm);
contextDocIdSet.and(surfaceFormDocIdSet);
return contextDocIdSet.cardinality();
}
/**
* Returns the number of documents containing the all terms in the set provided
* @param reader
* @param terms
* @return
* @throws java.io.IOException
*/
public long cardinality(IndexReader reader, Set<Term> terms) throws IOException {
long c = 0;
OpenBitSet result = null;
for (Term t: terms) {
OpenBitSet bs = (OpenBitSet) getDocIdSet(reader, t);
if (result==null)
result = bs;
else
result.and(bs);
}
if (result!=null) c = result.cardinality();
return c;
}
public long getMaxCacheSize() {
return mMaxCacheSize;
}
public void setMaxCacheSize(long maxCacheSize) {
this.mMaxCacheSize = maxCacheSize;
LOG.info(String.format("Setting the SurrogateCache.maxCacheSize to %s", maxCacheSize));
}
/**
* How many times does a surface form occur in the context field?
* For phrases, only approximation is possible.
* @param reader
* @param sf
* @return
*/
public long getPromiscuity(IndexReader reader, SurfaceForm sf) {
long p = 0;
Query q;
try {
// Use query parser to get analyzed terms
q = mLuceneManager.getQuery(new Text(sf.name()));
Set<Term> terms = new HashSet<Term>();
q.extractTerms(terms);
LOG.info(String.format("Terms: %s",terms));
// Now get how many documents contain all of those terms
p = cardinality(reader, terms);
} catch (Exception e) {
e.printStackTrace(); //To change body of catch statement use File | Settings | File Templates.
}
return p;
}
}